Unsupervised feature selection by non-convex regularized self-representation

作者：

Highlights：

• An ℓ2,1-2 self-representation unsupervised feature selection is proposed.

• ℓ2,1-2 is proved to guarantee the sparsity of selection matrix in theory.

• An iterative CCCP algorithm is designed to tackle the nonconvexity of ℓ2,1-2.

• The global convergence of our CCCP is theoretically analyzed.

• Extensive experimental results verify the effectiveness of the proposed method.

摘要

•An ℓ2,1-2 self-representation unsupervised feature selection is proposed.•ℓ2,1-2 is proved to guarantee the sparsity of selection matrix in theory.•An iterative CCCP algorithm is designed to tackle the nonconvexity of ℓ2,1-2.•The global convergence of our CCCP is theoretically analyzed.•Extensive experimental results verify the effectiveness of the proposed method.

论文关键词：Unsupervised feature selection,Self-representation,Non-convex regularization,CCCP,ADMM

论文评审过程：Received 21 May 2019, Revised 28 November 2020, Accepted 19 January 2021, Available online 29 January 2021, Version of Record 19 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114643